SMBMIP
SMBMIP
over the Greenland ice sheet
over the Greenland ice sheet
Intercomparion of 11 models
Intercomparion of 11 models
over 1980-2012 forced by ERA-Interim
over 1980-2012 forced by ERA-Interim
Model Name
Model type
Reso.
Institute (Country)
Contacts
EBM GCM PDD RCM
(km)
BESSI
x
10
Bergen Univ.
NO
Zolles T.
BOX13 (MM5)
x
5
GEUS
DN
Box J.
CESM
x
~1 (100?)
Utrecht Univ.
NL
Kampenhout L., Lenaerts J., …
dEBM
x
1
AWI
DE
Krebs-Kanzow U.
HIRHAM
x
5
x
DMI
DN
Mottram R., Langen P., …
MAR (ISMIP6)
x
15
Liège Univ.
BE
Fettweis X.
MPI-ESM
x
~1 (150)
Max Planck Inst.
DE
Kapsch M.
PDD1km
x
1
Sheffield Univ.
UK
Wilton D., Hanna E.
PDD5km
x
5
Lincoln Univ.
UK
Hanna E., Huybrechts P.
RACMO
x
~1 (5)
x
Utrecht. Univ
NL
Noel B., van den Broeke M., van de Berg W.J., …
SNOWMODEL
x
5
Colorado Univ.
US
Liston G., Mernild S.
Modis
Albedo
1. List of models
1. List of models
●
All the models have been interpolated on the same
ISMIP6 based 1km grid
by using the
downscaling technique of Franco et al. (2012), taking into account the sub-grid 1km topo.
●
The comparison is done on the
"common" ice sheet mask
and common period (
1980-2012
).
Mean snowline over 2000-2012 based on MODIS
(source: jonathan_ryan@brown.edu)
2. Mean 2000-2012 SMB
2. Mean 2000-2012 SMB
2. Mean 2000-2012 SMB
2. Mean 2000-2012 SMB
A priori, the mean snowline
should be in the ablation zone !!
Average of the 1980-2012 SMB
over the 11 models
Standard deviation of the mean
1980-2012 SMB over the 11 models
"Dark zone" only in a few
MODIS forced models
Ablation zone or accumulation zone
in the models
La
rg
e
di
sc
re
pa
nc
ie
s
in
th
e
pr
ec
ip
ita
tio
n
3. Ensemble mean
3. Ensemble mean
Very narrow
Models vs ensemble mean
Models vs ensemble mean
Units: m WE/yr
3. Ensemble mean
3. Ensemble mean
Model Bias (mWE) RMSE Correl. BESSI_zolles -0.07 0.11 0.87 BOX13_box 0.02 0.13 0.83 CESM_kampenhout 0.05 0.14 0.80 dEBM_krebs -0.01 0.08 0.89 HIRHAM_mottram 0.02 0.13 0.82 MAR_fettweis 0.01 0.08 0.93 MPI-ESM_kapsch 0.01 0.12 0.76 PDD1km_wilton -0.01 0.04 0.97 PDD5km_hanna -0.01 0.04 0.96 RACMO_noel -0.02 0.08 0.88 SNOWMODEL_liston -0.05 0.12 0.87 BESSI_zolles 0.37 0.87 0.79 BOX13_box -0.01 0.61 0.88 CESM_kampenhout 0.11 0.66 0.86 dEBM_krebs 0.03 0.70 0.84 HIRHAM_mottram 0.13 0.58 0.90 MAR_fettweis 0.10 0.49 0.93 MPI-ESM_kapsch -0.06 0.75 0.81 RACMO_noel -0.07 0.61 0.89 SNOWMODEL_liston -0.30 0.67 0.89 BESSI_zolles 0.45 0.89 0.81 BOX13_box -0.01 0.58 0.90 CESM_kampenhout 0.11 0.61 0.89 dEBM_krebs 0.05 0.76 0.82 HIRHAM_mottram 0.09 0.57 0.91 MAR_fettweis 0.10 0.48 0.93 MPI-ESM_kapsch -0.05 0.69 0.85 PDD1km_wilton -0.18 0.69 0.89 PDD5km_hanna -0.17 0.72 0.86 RACMO_noel -0.05 0.62 0.90 SNOWMODEL_liston -0.31 0.61 0.92
Ic
e
c
o
re
s
(
#
2
6
0
)
F
u
ll
P
R
O
M
IC
E
S
M
B
d
a
ta
B
a
se
(#
1
6
9
4
)
O
nl
y
m
ai
n
ic
e
sh
ee
t
S
M
B
d
at
a
ba
se
(#
14
39
)
Observations are not representative of the whole ice sheet !!
Observations are not representative of the whole ice sheet !!
Ice cores (Bales et al., ...)
PROMICE SMB data base
(Machguth et al., 2016)
Mean 1980-2012 SMB from MAR – 15km (mmWE/yr)
4. Evaluation
4. Evaluation
Percentage of the main ice sheet
Total Error
BOX13_box
1.0
2.3
3.3
dEBM_krebs
1.6
2.8
4.4
MAR_fettweis
0.5
2.5
2.9
CESM_kampenhout
1.5
0.9
2.4
BESSI_zolles
3.3
0.8
4.1
PDD1km_wilton
1.6
1.7
3.3
PDD5km_hanna
1.2
4.2
5.4
RACMO_noel
0.9
2.1
3.0
SNOWMODEL_liston
0.4
9.7
10.1
HIRHAM_mottram
0.9
1.6
2.5
MPI-ESM_kapsch
0.6
1.5
2.1
SMB > 200mm/yr
and bare ice in
MODIS
SMB < -200mm/yr
and no bare ice in
MODIS
Comparison of the mean 2000-2012 bare ice area from MODIS
Comparison of the mean 2000-2012 bare ice area from MODIS
with respect to the mean 2000-2012 SMB from models.
with respect to the mean 2000-2012 SMB from models.
In the bare ice area,
SMB should be negative!
SMB can be negative even
if bare ice is not detected by MODIS
4. Evaluation
4. Evaluation
Mean SMB vs Bare ice area
5. SMB vs GRACE
5. SMB vs GRACE
1
2
3
4
5
6
7
8
1
1
2
2
3
3
4
4
5
5
6
6
7
7
8
8
GrIS
GrIS
Source GRACE:
andreas.groh@tu-dresden.de
on the common
ice sheet mask
1
2
3
4
5
6
7
8
GrISGrIS Trend. R^2 RMSE
BESSI -143.8 0.98 86.9 BOX13 -83.6 0.98 79.7 CESM -188.1 0.98 94.9 dEBM -41.2 0.98 79.1 HIRHAM -81.1 0.99 51.9 MAR -73.6 0.99 48.4 MPI-ESM -177.2 0.93 186.5 PDD1km -91.4 0.98 82.6 PDD5km -80.5 0.98 74.6 RACMO -85.8 0.99 51.8 SNOWMODEL 44.2 0.98 90 Basin 2
Basin 2 Trend. R^2 RMSE
BESSI 1.3 0.72 8.4 BOX13 2.8 0.77 7.4 CESM 3.5 0.6 10.2 dEBM 13.3 0.63 12.3 HIRHAM 17.4 0.63 9.9 MAR 5.4 0.77 7.4 MPI-ESM -0.5 0.29 13.3 PDD1km 5.2 0.71 8.8 PDD5km 8.1 0.69 10.8 RACMO 7.7 0.74 7.8 SNOWMODEL 25.5 0.5 21.9 Basin 5
Basin 5 Trend. R^2 RMSE
BESSI -11.1 0.94 16.2 BOX13 -9.9 0.97 12.2 CESM -17 0.96 14.2 dEBM -2.6 0.95 14.3 HIRHAM -2.9 0.93 17.4 MAR -8 0.98 9.9 MPI-ESM -10.8 0.92 19.6 PDD1km -6.7 0.95 14.7 PDD5km -8.8 0.95 14.4 RACMO -13.1 0.97 11.2 SNOWMODEL -0.1 0.97 11.5
Basin 6Basin 6 Trend. R^2 RMSE
BESSI -10.5 0.89 36 BOX13 13.5 0.95 24.8 CESM -11.8 0.87 39.6 dEBM 17.8 0.93 27.9 HIRHAM -11.8 0.79 49.6 MAR 9.8 0.97 18.9 MPI-ESM -10.9 0.65 66.3 PDD1km 1.5 0.92 31.9 PDD5km 4.5 0.92 30.5 RACMO 2.9 0.97 18.3 SNOWMODEL 27.6 0.92 29.7
